search

hbc / bcbioRNASeq

R package for bcbio RNA-seq analysis.
https://bioinformatics.sph.harvard.edu/bcbioRNASeq
GNU Affero General Public License v3.0
58 stars 21 forks source link

Use non-normalized counts by default for gene saturation methods. #107

Closed roryk closed 6 years ago

mjsteinbaugh commented 6 years ago

@roryk The code inside the plotGeneSaturation() function is using match.arg() to capture the first argument in the vector by default. I don't think we can mix logicals (e.g. FALSE) with characters (e.g. "tpm") here unless we change the match.arg() call. It might be simpler to just set normalized = FALSE by default, and then allow tpm or tmm as alternates. Or alternatively, should we just always use the raw counts here? That would be a simpler approach.

roryk commented 6 years ago

Ah, gotcha. Always raw counts makes sense to me, I can swap to that.

mjsteinbaugh commented 6 years ago

Yeah I think that's the way to go. Should we include a low pass filter to remove genes with very low expression that aren't really "detected"?

mjsteinbaugh commented 6 years ago

Nice – before merging the pull request run devtools::document(), which will update the plotGeneSaturation.Rd file in the man directory. Then it should be good to go.

mjsteinbaugh commented 6 years ago

Ah never mind, I see that's merging into develop. I'll add that to the update